Disulfiram Without Reliable Supervision: When a Deterrent Drug Needs a Witness
A patient specifically requests disulfiram and can explain exactly why she wants it. What she can’t yet offer is anyone to confirm she takes it during the hours she’s actually at risk — and the drug’s own founding trial says that gap matters.
“I want something that makes the decision for me.” That's how K.D., forty-four, framed her own request when she brought up disulfiram at this visit — not a question so much as something she had already thought through and arrived at on her own before the appointment. She has never been through formal treatment, but her drinking has been a recognized problem for about four years, escalating clearly over the last one as her marriage deteriorated: a bottle of wine most nights, more on weekends, and two prior attempts to cut down on her own that both failed within weeks. She doesn't trust her willpower alone anymore, and says so without embarrassment — she is motivated, articulate about why she's asking, and has clearly thought the request through rather than arriving at it casually.
She moved into her sister's spare room three weeks ago after her apartment lease fell through in the middle of a separation she hadn't planned on. She works part-time doing alterations out of a shop two bus stops away, a job she has held for six years and genuinely likes, and has no other significant medical history — no liver disease, no cardiac history, no psychiatric diagnosis beyond the drinking itself. Her sister has offered to help however she can, but works the overnight shift at a distribution center three nights a week and cannot simply change that schedule on short notice. Those are exactly the hours, K.D. admits, when she used to drink most — and for now, on those particular nights, there is no one else in the apartment who could reliably watch her take a dose.
A patient asks for the drug that requires someone watching her take it
She asked for this by name, described the mechanism accurately, and told me directly why she wants it: she doesn't trust herself alone with a choice, and she wants a drug that removes the choice. That's not a naive request — that's someone who has tried twice on willpower alone and correctly diagnosed what didn't work. I don't think we should substitute our judgment about supervision for her own honest account of what she needs.
I want to be precise about what the evidence actually supports, because it's easy to conflate "disulfiram works" with "disulfiram works the way she's about to take it." The VA Cooperative Study — Fuller and colleagues, 1986, still the largest controlled trial of disulfiram ever run — randomized 605 men to 250 mg, to a 1 mg dose that preserved the threat of a reaction, or to no disulfiram at all, and found no significant difference among the three in total abstinence or time to first drink. The finding that actually matters for her sits underneath that: adherence predicted abstinence in every arm, including the inert 1 mg dose. Taken unsupervised, that trial simply can't show the drug is doing the work — the disulfiram literature that does show a benefit is overwhelmingly the open-label, observed-dosing literature.
Her request is genuine, but genuine motivation and confirmed daily supervision are two different things — and Fuller's own data are exactly why we can't let one stand in for the other, since the men who did best were the ones who reliably took their dose regardless of which arm they were in.
You're right that the efficacy data specifically requires supervision, and I don't think that point is answerable on her motivation alone. But I want to add the piece neither of you has named: the disulfiram-ethanol reaction — flushing, tachycardia, nausea, and in rarer cases hypotension severe enough to be dangerous — isn't just a deterrent story, it's a real physiologic event, and I'm not comfortable prescribing that risk into hours where she's alone with no one to call for her if it happens badly.
What I'd propose instead: build supervision before we build the prescription. If her sister can shift even one or two of her night shifts, or if K.D. is willing to do observed dosing through a pharmacy or a structured outpatient program three days a week to start, disulfiram becomes a real option rather than an unsupervised gamble on both the efficacy and the safety questions at once.
Agreed: naltrexone started today; K.D. to speak with her sister about shifting one night shift and to check whether her regular pharmacy offers observed dosing three days a week; disulfiram reconsidered once one of those is confirmed, not indefinitely deferred.
Not agreed: whether disulfiram should ever be offered to a patient in her situation without a guaranteed supervision plan already in place, or whether a strongly motivated, well-informed request like hers is itself enough reason to start now and build supervision around it. The addiction psychiatrist would have started today; the addiction medicine specialist and primary care physician wanted the structural piece in place first. K.D. left the visit siding with waiting, on her own reasoning that a reaction happening while she's alone would set her back further than a few more weeks would.